AMD Athlon 64 X2 running very HOT!!! SOLVED!

[WildViper]

Originally posted by: Eeqmcsq
You'll have to try the finger-on-the-heatsink test to see if it really gets that hot. Otherwise, I think your temps are reading off bad Brisbane sensors.

Ok, did this one. I touched both sides of the heatsink(not cooper) and ran the FLV test again. The temps went up, but the heatsink remained ok to touch. Slight warmth, but nothing scorching!

So how do I fix this bad sensor than? I can't process anything cause the system shutsdown. This is retarded...AMD should issue new processors to us.

Is an upgrade the only way to fix this? If so, what can I get for my Gigabyte GA-MA78GM S2H board? It says it is AM3 ready(not sure what that means).

I don't wanna spend too much money..less than a hundred(or even $70). I feel cheated by AMD. They should stand behind their products and fix these "known" issues.

 

[Eeqmcsq]

Sadly, there is no way to fix it. But you should turn off your BIOS auto shut down feature to make sure your shut down is really caused by the BIOS. I had that issue with my Pentium 4 where it would suddenly shut down, usually after a prolonged game of Starcraft, and I thought it was a temp issue. But then the shut downs occurred randomly, even after just cold booting and logging in. Turns out my power supply was dying. I replaced it and the shut downs went away.

 

[DominionSeraph]

I'd still say your heat sink is only making partial contact. It being a bad sensor dosen't jive with this being a recent issue, with it scaling with load, nor with the auto-shutdown.

 

[Grabo]

I would say the sensor is shot. Haven't seen a CPU @ 107C but I imagine it would be experiencing severe issues then. The fact that temps drop instantly would also suggest this.
If you know the sink makes contact with the CPU (because of where the thermal paste has stuck to both surfaces when you remove it) and the case doesn't go super hot (I hardly believe that) then just turn off the BIOS CPU temp shutdown condition, as mentioned by Eeqmcsq. (Such a condition is the only thing that would make the machine shut down non-permanently as far as I'm aware)
(If the thing does eventually overheat and catch fire and burn down an entire office then..that's just bad luck )

 

[imported_Asianman]

This sounds like on of Thermaltake's AMD only coolers. (non copper, 2500 fan RPM) Because if it is, then change it immediately. I have a "TMG A3" on a X2 5000 and it immediately shuts down when I'm trying to play games. Have you tried the stock cooler? Changing back to the stock cooler immediately brought the temps down to about 50C

 

[WildViper]

Originally posted by: DominionSeraph
I'd still say your heat sink is only making partial contact. It being a bad sensor dosen't jive with this being a recent issue, with it scaling with load, nor with the auto-shutdown.

I have been thinking about this problem(obviously) and when I have seen it before. Well, it seems that I have seen this issue before when I used the computer to do some actual work!!!

I just didn't recall since I have not been using this comp, my dad has. He uses it for surfing and very basic stuff. Only lately I have been using it and will continue now on.

Also, the heatsink is making full contact...verified by seeing how the paste was stuck from previously and once again today after I applied the new Ceramique. I re-did this again just to make sure. No changes.

This sounds like on of Thermaltake's AMD only coolers. (non copper, 2500 fan RPM)

It is. This is it: Thermaltake CL-P0444

I will try and switch out the fan tomorrow. I am not sure where I have thrown the stock cooler!!! Will buy a new one.

By the way, thanx all for helping me out.

 

[WildViper]

Originally posted by: Grabo
I would say the sensor is shot. Haven't seen a CPU @ 107C but I imagine it would be experiencing severe issues then. The fact that temps drop instantly would also suggest this.
If you know the sink makes contact with the CPU (because of where the thermal paste has stuck to both surfaces when you remove it) and the case doesn't go super hot (I hardly believe that) then just turn off the BIOS CPU temp shutdown condition, as mentioned by Eeqmcsq. (Such a condition is the only thing that would make the machine shut down non-permanently as far as I'm aware)
(If the thing does eventually overheat and catch fire and burn down an entire office then..that's just bad luck )

One thing that I did notice...At somepoint when my system was auto shutting down, I checked the Bios and I had the Temp Warning Off.

I will repeat this and see what happens.

 

[Eeqmcsq]

One thing that I did notice...At somepoint when my system was auto shutting down, I checked the Bios and I had the Temp Warning Off..

Wait a sec. Describe what kind of shut down you see. Is it a clean WIndows shut down that goes through the normal WIndows shut down sequence, or is it a sudden shut down like when you press the power button?

 

[porksandwich9113]

Brisbane core tends to run alot hotter I might add.
My windsor runs at 47 C tops (AC5, and a Scythe SCNJ HSF - Clocked at 3.2ghz)
My friends Brisbane hits like 58 C /w AC5, and a Cool Master Hyper-something, clocked at 3.2ghz)

Also, AMD processors are pretty cheap.
You can pick up a nice dual core for 60, and I am sure someone would buy that brisbane that has a sensor malfunction.

 

[DominionSeraph]

Originally posted by: Grabo
The fact that temps drop instantly would also suggest this.

But not exclusively.
I've run with no paste at all. I'm currently running with butter. (I Can't Believe It's Not Thermal Compound.)
Temps drop instantly and dramatically in both scenarios when transitioning to idle.

 

[Lemon law]

Does it really matter, even if cpu-z will effortlessly identify your processor. We are talking 95 watts here, far too much evident heat, and if the system shuts down even if the sensors lie, the system still shuts down. And even if you have lying sensors, having fans than rev up louder than 10 sewing machines right in your ear, hardly makes computers a pleasant experience.

I have the same problem with my wife's intel pentium 915 D pressler, and though its taken over a year to get to the critical point where she gets blue screens of death and system shut downs, I am simply forcing her to bite the bullet and get a 65 watt e5200 instead. And I will bet she will gain more speed, stability, and quiet in the process.

I for one say piss on all 95+ watt processors, because pissing on them may be the only way to keep them cool enough.

I have been running my e5200 all day, and have yet to exceed 42C. Brand new, my wife's 915 idled at 50C, and now I have to underclock to get it to idle at 66 C. I too tried the remount the processor trick with no effect. Maybe my wife got a defective processor, does it matter when intel will not stand behind it, I just want an end to grief. And 95+ watts is always the wrong road.

 

[WildViper]

Originally posted by: Eeqmcsq

One thing that I did notice...At somepoint when my system was auto shutting down, I checked the Bios and I had the Temp Warning Off..

Wait a sec. Describe what kind of shut down you see. Is it a clean WIndows shut down that goes through the normal WIndows shut down sequence, or is it a sudden shut down like when you press the power button?

It is a abrupt shutdown. One second it is on, the next it is dead.

Well, from the discussions here, rather than buying a new heatsink, I think the better route is to buy a new CPU that is less than 95watts. No matter what, it seems the Brisbanes run hot. I want peace of mind and that I can actually do some heavy duty tasks without worrying my house will burn down!!!

I will try and sell that one.

Anyways, can you guys/gals recommend a good AMD AM2/AM3 processor for less than $100 that doesn't have this stupid heat issue?

 

[VirtualLarry]

Originally posted by: WildViper
Anyways, can you guys/gals recommend a good AMD AM2/AM3 processor for less than $100 that doesn't have this stupid heat issue?

See the threads in this forum about the Athlon II X4 CPU for $99. Looks like a winner!

 

[Eeqmcsq]

If your BIOS auto shut down was turned off, and your finger didn't get burned by the heatsink, and your PC abruptly shut down out of the blue, I'm leaning more in favor of a PSU failure right now. That's based on the one time that it happened to me.

 

[WildViper]

Originally posted by: Eeqmcsq
If your BIOS auto shut down was turned off, and your finger didn't get burned by the heatsink, and your PC abruptly shut down out of the blue, I'm leaning more in favor of a PSU failure right now. That's based on the one time that it happened to me.

No, no, no, no...don't tell me that. Please no!!

Dammit, I was almost going to go to sleep peacefully knowing that I just had to swap out the CPU and all what ailes me will be gone!!!! With your new direction, it will never end.

What I don't get is how is it the PSU if all what you say :

If your BIOS auto shut down was turned off, and your finger didn't get burned by the heatsink, and your PC abruptly shut down out of the blue

is true. Can a PSU cause the CPU temps to jump that high? Or are you saying that it is a bad sensor combined with a faulty PSU?

Well, how come the system never shuts downs while doing normal everyday stuff? It only shuts downs when I am doing something more muscle flexing...like exporting to Jpegs, of converting FLV to MP4 and only when the temps are getting high to 100C.

I still have to run the bios test again to ensure that it is going off without that warning turned on.

 

[DominionSeraph]

Originally posted by: Eeqmcsq
and your finger didn't get burned by the heatsink,

Your finger would only get burned by the heatsink if your processor was on fire. I run 60c Prime95 and my heat sink barely gets warm.

And it's not an overloaded PSU. That would trip on current load, not core temps.

Wildviper, keep playing with the HSF. Once it took me 5 tries just to get my system to POST, even though it looked by the thermal pad that everything was making contact. (And it was the heat sink causing the problem. As soon as I got it down good everything was fine.)

 

[bryanW1995]

Originally posted by: WildViper

Originally posted by: Eeqmcsq
If your BIOS auto shut down was turned off, and your finger didn't get burned by the heatsink, and your PC abruptly shut down out of the blue, I'm leaning more in favor of a PSU failure right now. That's based on the one time that it happened to me.

No, no, no, no...don't tell me that. Please no!!

Dammit, I was almost going to go to sleep peacefully knowing that I just had to swap out the CPU and all what ailes me will be gone!!!! With your new direction, it will never end.

What I don't get is how is it the PSU if all what you say :

If your BIOS auto shut down was turned off, and your finger didn't get burned by the heatsink, and your PC abruptly shut down out of the blue

is true. Can a PSU cause the CPU temps to jump that high? Or are you saying that it is a bad sensor combined with a faulty PSU?

Well, how come the system never shuts downs while doing normal everyday stuff? It only shuts downs when I am doing something more muscle flexing...like exporting to Jpegs, of converting FLV to MP4 and only when the temps are getting high to 100C.

I still have to run the bios test again to ensure that it is going off without that warning turned on.

It could be your psu, what psu do you have?

However, I think that it is probably just a poorly seated heatsink/fan. keep trying to reseat it. first, COMPLETELY take off all the thermal paste with isopropyl alcohol. then try a very small amount (think piece of rice) in the exact center. If that doesn't work add a little more, use a credit card to spread out evenly. If you continue along this route 5-6 times then you could try buying new parts, just keep in mind that this could be the cpu, cpu cooler, psu, or some other random event. I've thrown a lot of money in the past at issues like this and eventually gave up and bought a new system.

 

[Eeqmcsq]

Dammit, I was almost going to go to sleep peacefully knowing that I just had to swap out the CPU and all what ailes me will be gone!!!! With your new direction, it will never end

Sorry, but I have to throw it out there as a possibility because of first hand experience. If you can swap CPUs and the problem goes away, then it really was the CPU. If not, then my next guess is the PSU.

Can a PSU cause the CPU temps to jump that high? Or are you saying that it is a bad sensor combined with a faulty PSU?

I'm saying not to trust your temp sensors. Before this mysterious shut down problem started happening, were you monitoring your CPUs temps back then? Were they more reasonable, even at load? Or did you only start monitoring your CPU temps when this problem started?

Well, how come the system never shuts downs while doing normal everyday stuff? It only shuts downs when I am doing something more muscle flexing...like exporting to Jpegs, of converting FLV to MP4 and only when the temps are getting high to 100C.

Because your CPU intensive tasks draw more power from the PSU, and if the PSU can't supply that power any more, the PC will choke. That's why I say the PSU is another good candidate for the problem.

Your finger would only get burned by the heatsink if your processor was on fire. I run 60c Prime95 and my heat sink barely gets warm.

Yes, I messed up. What I should have said was to put the finger on the heatsink, both at idle and at load. If the CPU is REALLY approaching 100C as the sensors claim, and the heatsink was properly seated, the heatsink should be venting off a noticeable amount of heat.

 

[DominionSeraph]

Originally posted by: Eeqmcsq
If the CPU is REALLY approaching 100C as the sensors claim, and the heatsink was properly seated, the heatsink should be venting off a noticeable amount of heat.

Yes, but that's not really a viable combination. His processor would have to be drawing like 400w to get to 100c with a properly mounted HSF. There's no way it's doing that in a condition that's still operational.

If the temp readings are correct, it's not getting full cooling. Plain and simple.

 

[WildViper]

Ok, just an update. I reseated the heatsink again today. Ran the tests and same exact results.

Dammit, I forgot to run the bios test. Will post back soon here the results from that.

 

[BOLt]

I also have a very hot A64 X2 3600+ Brisbane clocked to 2800-2900 MHz. My temps hit ~80C during gaming (though it's more than likely due to poor ventilation and a lot more equipment than my case is suited for). I suspect I'll kill the processor and/or motherboard before too long, but at over 2 years of 24/7 operation I guess I can look to hope for another 9 months out of her.

I'm up for trying anything to get those temps down though (other than dropping the speed), so if the OP comes up with some fantastic solution to his problem, he would be greatly appreciated in sharing this with me!

:cheers:

 

[HannibalX]

It's that piece of crap heatsinc/fan you're using. Get rid of it ASAP. It can't cool that CPU. It doesn't even appear to be half an inch thick, no copper, no heat pipes (stock AMD heatsinc has heat pipes).

I have this unit: http://www.newegg.com/Product/...E16835887002&Tpk=BTF90

My Opteron 185 (which is essentially an Athlon 64 X2) idles at 30C with this unit and never hits higher than 45C loaded (at stock 2.6Ghz).

 

[WildViper]

Ok everyone...ready for this??? New Record!!!!!!!

I hit 129C on Core 1 and 100C on Core 0!!!!!!!!!!!!!!!!!!!

Yes, that is C and I do know the difference between C and F.

Unfortunately I couldn't capture the HWInfo's screen cause the system shutdown. By the way, I have disabled the Bios shutdown/warning thingy and it STILL shut down.

Here is a picture of something that I did capture before it got to 129C. 123C!

So what have I done now??

• Reseated the heatsink..probably about 5 times now. A PITA I tell you, especially cleaning and reapplying the paste. At this rate, I am going to run out of paste.

• Turned off the Bios warning for CPU temps to see if it will shut down. It did

• Cleaned everything...It is so clean, I can probably put my food on it and eat!

• I have bought a new Heatsink..Coolermaster Vortex 752 but will install it later today

Again, thanx for everyone's help. I am trying everything as you can tell.

Originally posted by: iFX
It's that piece of crap heatsinc/fan you're using. Get rid of it ASAP. It can't cool that CPU. It doesn't even appear to be half an inch thick, no copper, no heat pipes (stock AMD heatsinc has heat pipes).

 

[HannibalX]

The Vortex 752 should offer similar cooling to the OEM AMD cooler, which will be mid high 40s to low 50s under load and mid 30s at idle. Good luck!

 

[HannibalX]

Alright, so what's the outcome? LOL